AI projects

start trial and erroring different settings on different systems to make musicgen run properly on the GPU

2024-01-14 06:48:36 UTC

Shashank

spend time learning about pytorch and CUDA to speed up generation times

2024-01-12 05:26:23 UTC

Shashank

make community posts asking for advice on speeding up generation time

2024-01-11 07:21:31 UTC

Shashank

continue trying to improve generation time

2024-01-10 06:46:11 UTC

Shashank

try 3 different instances to improve musicgen generation time

2024-01-09 06:30:42 UTC

Shashank

try bringing down musicgen generation time

2024-01-08 07:30:31 UTC

Shashank

redeploy sagemaker endpoint

2023-12-30 06:30:31 UTC

Shashank

continue readings

2023-12-29 06:31:43 UTC

Shashank

read NLP chapter of fast ai book

2023-12-28 07:28:34 UTC

Shashank

set up sagemaker endpoint

2023-12-27 07:14:04 UTC

Shashank

study stemgen

2023-12-21 03:15:32 UTC

Shashank

continue experimenting with different models

2023-12-20 06:01:40 UTC

Shashank

experiment with a series of AWS models to find the best and cheapest

2023-12-19 06:34:47 UTC

Shashank

fix short duration issue for audio files

2023-12-18 09:06:04 UTC

Shashank

Try new things to get audio gen to 30 seconds

2023-12-16 04:16:45 UTC

Shashank

read about transformers

2023-12-15 08:02:17 UTC

Shashank

keep learning fastai

2023-12-14 07:25:24 UTC

Shashank

Make forum post on hugging face explaining issue

2023-12-13 02:17:02 UTC

Shashank

try older model version and continue experimenting with optimizations